JP7163640B2 - Synchronous motor controller - Google Patents

Description

The present invention relates to autotuning performed by a control device for a synchronous motor.

A method of optimizing a magnetic flux model considering magnetic saturation including interference between d and q axes by auto-tuning in an embedded permanent magnet type synchronous motor (IPMSM) is known. As an example, a parameter corresponding to the maximum value of the slope of the q-axis magnetic flux with respect to the q-axis current, a parameter indicating the degree of magnetic saturation on the q-axis, and a parameter indicating the degree of interference of the d-axis current with the q-axis magnetic flux. There is a method of auto-tuning (see Patent Document 1, for example).

The magnetic saturation characteristic is the characteristic that the linearity between the d- and q-axis magnetic fluxes and the respective axis currents corresponding to them is lost due to the magnetic saturation of the motor iron core as the current increases. A characteristic is that the self-axis magnetic flux changes due to the influence of the other-axis current.

JP 2015-144502 A

In order to control the torque of a synchronous motor such as an IPMSM or a synchronous reluctance motor (hereinafter referred to as SynRM) with high accuracy, a magnetic flux model that takes into account at least the magnetic saturation characteristics of the motor core is obtained, and the current is controlled based on this magnetic flux model. It is desirable to In order to perform current control based on the magnetic flux model, it is necessary to calculate each parameter of the magnetic flux model.

However, in the auto-tuning method that requires no-load operation of the synchronous motor in order to calculate each parameter of the magnetic flux model, auto-tuning cannot be performed when the synchronous motor is connected to a load device.

Accordingly, the present disclosure provides a control device for a synchronous motor that can perform auto-tuning even when the synchronous motor is connected to a load device.

As one aspect of the technology of the present disclosure,
The current and voltage supplied to the synchronous motor by the power converter are controlled on a d, q orthogonal rotating coordinate system consisting of a d-axis parallel to the magnetic pole direction of the rotor of the synchronous motor and a q-axis orthogonal to the d-axis. a controller,
A computing device that performs computation based on a magnetic flux model that takes into account at least the magnetic saturation characteristics of an iron core of the motor,
The computing device is
controlling the DC component of the d-axis current of the synchronous motor to a command value;
applying a sinusoidal alternating voltage to the d-axis;
calculating the Fourier coefficient of the fundamental wave component of the d-axis current;
calculating the Fourier coefficient of the double harmonic component of the d-axis current;
calculating the Fourier coefficient of the triple harmonic component of the d-axis current;
calculating the Fourier coefficient of the fundamental wave component of the d-axis voltage of the synchronous motor;
calculating the Fourier coefficient of the fundamental wave component of the d-axis magnetic flux of the synchronous motor based on the Fourier coefficient of the fundamental wave component of the d-axis voltage and the angular frequency of the alternating voltage;
Fourier coefficient of the fundamental wave component of the d-axis current, Fourier coefficient of the double harmonic component of the d-axis current, Fourier coefficient of the triple harmonic component of the d-axis current, Fourier coefficient of the DC component of the d-axis current , and a first parameter indicating the maximum value of the slope of the d-axis magnetic flux with respect to the d-axis current, based on the Fourier coefficient of the fundamental wave component of the d-axis magnetic flux, and a second parameter indicating the degree of magnetic saturation on the d-axis 2 parameters,
A control device for a synchronous motor is provided, wherein the magnetic flux model is configured using the first parameter and the second parameter.

Further, as one aspect of the technology of the present disclosure,
The current and voltage supplied to the synchronous motor by the power converter are controlled on a d, q orthogonal rotating coordinate system consisting of a d-axis parallel to the magnetic pole direction of the rotor of the synchronous motor and a q-axis orthogonal to the d-axis. a controller,
A computing device that performs computation based on a magnetic flux model that takes into account at least the magnetic saturation characteristics of an iron core of the motor,
The computing device is
controlling the DC component of the d-axis current of the synchronous motor to a command value;
applying a sinusoidal alternating voltage to the q-axis;
calculating the Fourier coefficient of the fundamental wave component of the q-axis current of the synchronous motor;
calculating the Fourier coefficient of the triple harmonic component of the q-axis current;
calculating the Fourier coefficient of the DC component of the d-axis current;
calculating the Fourier coefficient of the double harmonic component of the d-axis current;
calculating the Fourier coefficient of the DC component of the d-axis magnetic flux of the synchronous motor based on the Fourier coefficient of the DC component of the d-axis current and the Fourier coefficient of the double harmonic component of the d-axis current;
Fourier coefficient of the fundamental wave component of the q-axis current, Fourier coefficient of the triple harmonic component of the q-axis current, Fourier coefficient of the DC component of the d-axis current, Fourier coefficient of the double harmonic component of the d-axis current , based on the Fourier coefficient of the DC component of the d-axis magnetic flux, a first parameter indicating the maximum value of the slope of the d-axis magnetic flux with respect to the d-axis current, and a second parameter indicating the degree of magnetic saturation on the d-axis , calculating a third parameter indicating the degree of interference of the q-axis current with the d-axis magnetic flux,
A control device for a synchronous motor is provided, wherein the magnetic flux model is constructed using the first parameter, the second parameter and the third parameter.

Further, as one aspect of the technology of the present disclosure,
The current and voltage supplied to the synchronous motor by the power converter are controlled on a d, q orthogonal rotating coordinate system consisting of a d-axis parallel to the magnetic pole direction of the rotor of the synchronous motor and a q-axis orthogonal to the d-axis. a controller,
A computing device that performs computation based on a magnetic flux model that takes into account at least the magnetic saturation characteristics of an iron core of the motor,
The computing device is
controlling the DC component of the d-axis current of the synchronous motor to a command value;
applying a sinusoidal alternating voltage to the q-axis;
calculating the Fourier coefficient of the DC component of the d-axis current;
calculating the Fourier coefficient of the DC component of the d-axis magnetic flux of the synchronous motor based on the Fourier coefficient of the DC component of the d-axis current and the Fourier coefficient of the double harmonic component of the d-axis current;
the d-axis current, the q-axis current of the synchronous motor, the Fourier coefficient of the DC component of the d-axis magnetic flux, a first parameter indicating the maximum value of the slope of the d-axis magnetic flux with respect to the d-axis current, and calculating a third parameter indicating the degree of interference of the q-axis current with the d-axis magnetic flux based on the second parameter indicating the degree of magnetic saturation;
A control device for a synchronous motor is provided, wherein the magnetic flux model is constructed using the first parameter, the second parameter and the third parameter.

According to the technology of the present disclosure, autotuning can be performed even when the synchronous motor is connected to a load device.

It is a block diagram showing an example of whole composition of a 1st embodiment. 4 is a first schematic example of the d-axis current flowing in the first embodiment; 4 is a second schematic example of the d-axis current flowing in the first embodiment; FIG. 3B is a third schematic example of the d-axis current flowing in the first embodiment; FIG. FIG. 11 is a block diagram showing an example of the overall configuration of second to fifth embodiments; 10A and 10B are schematic examples of the d-axis current and the q-axis current flowing in the second to fifth embodiments; FIG. 11 is a diagram illustrating sample points of instantaneous values used in the fourth and fifth embodiments; It is a figure which illustrates the hardware constitutions of the arithmetic unit with which a control apparatus is provided.

Hereinafter, embodiments of a synchronous motor control device according to the present disclosure will be described with reference to the drawings. The same reference numerals are given to the same components, and redundant explanations are omitted. It should be noted that the present invention is not limited to the following embodiments, and can be modified as appropriate without changing the gist of the invention.

The control device for a synchronous motor according to the present embodiment applies an alternating voltage while the rotor of the synchronous motor is stationary, and performs auto-tuning to calculate each parameter of the magnetic flux model from the response of the current that flows when the alternating voltage is applied. Further, the control device of this embodiment includes a parameter corresponding to the maximum value of the slope of the q-axis magnetic flux with respect to the q-axis current of the synchronous motor, a parameter indicating the degree of magnetic saturation on the q-axis, and a d-axis current with respect to the q-axis magnetic flux. Auto-tuning parameters that indicate the degree of interference. The control device of the present embodiment auto-tunes the magnetic flux model of the synchronous motor by applying a system identification method using a gradient vector based on a partial differential expression of each parameter during auto-tuning.

According to the control device of the present embodiment, no-load operation is not required when calculating each parameter of the magnetic flux model of the synchronous motor. It is possible to perform auto-tuning of the magnetic flux model in this state. Further, among the plurality of parameters of the magnetic flux model, a parameter indicating the maximum value of the slope of the q-axis magnetic flux with respect to the q-axis current, a parameter indicating the degree of magnetic saturation on the q-axis, and the degree of interference of the d-axis current with the q-axis magnetic flux A system identification method is applied in computing the parameters that indicate By applying the system identification method, auto-tuning can be performed regardless of the format of the magnetic flux model formula.

The control device of the present embodiment can control the rotation of the synchronous motor with high precision by using the magnetic flux model for which auto-tuning of each parameter has been completed for current control of the synchronous motor. For example, the control device uses a d-axis current command value i _d ^* and a q-axis current command value i _d ^* calculated based on a magnetic flux model in which each parameter after auto-tuning is reflected. It controls the current and voltage supplied to the synchronous motor. This makes it possible to control the rotation of the synchronous motor with high accuracy.

Next, the details of the control device of this embodiment will be described. First, the magnetic flux model used by the control device of this embodiment will be described.

<1. Magnetic flux model>
Equations 1 and 2 show the IPMSM magnetic flux model. Moreover, the magnetic flux model of SynRM is shown in Numerical formulas 3 and 4. These magnetic flux models take into account at least the magnetic saturation characteristics of the motor core. The SynRM magnetic flux model shown in Equations 3 and 4 is equivalent to Equations 1 and 2 with the equivalent magnetizing current _I0 and the magnetic flux offset φ0, which are parameters due to the permanent magnet, set to _zero . In the following embodiments, these models will be used as examples to explain how to calculate the parameters (K _Ld , K _sd , K _sdq , K _Lq , K _sq , K _sqd ) of the magnetic flux model.

<2. First Embodiment>
(2.1) Overall Configuration of First Embodiment FIG. 1 is a block diagram showing a control device according to the first embodiment together with a main circuit. or SM) voltage and current control method will be described together with the configuration of the control device. The control calculation of the voltage and current supplied to the synchronous motor by the power converter is performed on the d- and q-axis orthogonal rotating coordinates, and the magnetic pole (N-pole) direction of the rotor of the motor is the d-axis, A direction leading 90° from the d-axis parallel to the direction is defined as the q-axis.

The control device 100a of the present embodiment controls the DC component of the d-axis current to a command value, and controls the magnetic saturation characteristics of the motor iron core based on the current and voltage when a sinusoidal alternating voltage is applied to the d-axis. At least auto-tuning is performed to calculate each parameter of the considered magnetic flux model.

In FIG. 1, the integrator 5a integrates the angular frequency _ωh of the alternating voltage to calculate the angle _θh of the d-axis alternating current command value. The d-axis alternating current command calculator 6a calculates the AC component i _dh ^* of the d-axis alternating current command value as shown in Equation (5).

The adder 7a adds the d-axis current DC component command value _Idc(0) ^* and _idh ^* to calculate the _d -axis current command value id ^* . The d-axis current DC component command value _Idc(0) ^* may be set to zero.

The current coordinate converter 8a converts the u- and w-phase current detection values i _u and i _w respectively detected by the u-phase current detector 9ua and the w-phase current detector 9wa based on the magnetic pole position detection value θ ₁ of the synchronous motor 1a. coordinate conversion to the _d- and _q -axis current detection values id and iq.

The low-pass filter 10a removes high frequency components from the _d -axis current detection value id to calculate the d-axis current detection value _idf .

The deviation between the d-axis current command value i _d ^* and the d-axis current detection value i _df is calculated by the subtractor 11a, and this deviation is amplified by the d-axis current controller 12a to obtain the d-axis voltage command value v _dACR (d A shaft voltage feedback control value v _dACR ) is calculated. The d-axis current controller 12a operates so that the deviation between the d-axis current command value id ^* and the _d -axis current detection value _{idf becomes zero to calculate the d-axis voltage feedback control value vdACR .}

As will be described later, the parameters of the magnetic flux model are estimated using harmonic currents that flow due to magnetic saturation of the motor core. Therefore, in order to prevent the d-axis current adjuster 12a from acting on harmonic currents, the response frequency of the d-axis current adjuster 12a and the cutoff frequency of the low-pass filter 10a are adjusted to the angular frequency _ωh of the alternating voltage. It is set smaller than twice.

The DC armature resistance compensator 13a calculates the d-axis armature resistance feedforward compensation value v _dra by Equation (6).

The voltage compensation value calculator 14a uses the d-axis reactance estimated value _Xdhest and the d-axis armature resistance estimated value _Rdhest calculated by the impedance estimator 15a to calculate the d-axis voltage feedforward compensation value _{vdhFF according} to Equation 7. do. Estimation of impedance such as the d-axis reactance estimated value X _dhest and the d-axis armature resistance estimated value R _dhest may be performed by the impedance estimating unit 15a using a known technique described in Patent Document 1, for example. It is possible.

The Fourier coefficient calculator 16a calculates Fourier coefficients of the d-axis voltage and the _d -axis current based on _θh , _vd ^* , and id. The Fourier coefficient calculator 16a calculates the Fourier cosine coefficient _Vd(a1) of the fundamental wave component of the d-axis voltage, the Fourier sine coefficient _Vd(b1) of the fundamental wave component of the d-axis voltage, and the Fourier coefficient of the DC component of the d-axis current. I _d(0) , Fourier cosine coefficient I _d(a1) of the fundamental wave component of the d-axis current, Fourier sine coefficient I _d(b1) of the fundamental wave component of the d-axis current, Double harmonic component of the d-axis current Fourier cosine coefficient I _d(a2) , Fourier sine coefficient I _d(b2) of the second harmonic component of the d-axis current, Fourier cosine coefficient I _d(a3) of the third harmonic component of the d-axis current, and d Calculate the Fourier sine coefficient _Id(b3) of the triple harmonic component of the shaft current. Fourier coefficients can be calculated using a known technique described in Patent Document 1, for example.

The Fourier cosine coefficient I _d(a2) of the double harmonic component of the d-axis current, the Fourier sine coefficient I _d(b2) of the double harmonic component of the d-axis current, and the d-axis calculated by the Fourier coefficient calculator 16a. From the Fourier cosine coefficient I _d(a3) of the third harmonic component of the current and the Fourier sine coefficient I _d(b3) of the third harmonic component of the d-axis current, the armature resistance compensator 17a is calculated as Calculate the harmonic armature resistance compensation value _{v_dhra} .

The d-axis voltage command value v _d ^* is obtained by adders 7b and 7c to obtain a d-axis voltage feedback control value v _dACR , a d-axis armature resistance feedforward compensation value v _dra , a d-axis voltage feedforward compensation value v _dhFF , a d-axis It is calculated by adding the harmonic armature resistance compensation value _{v_dhra} . On the other hand, the q-axis voltage command value v _q ^* is fixed at zero.

The _d- and _q -axis voltage command values vd ^* and vq ^* calculated as described above are converted by the voltage coordinate converter 18a to the u-, v- and w-phase _phase voltage command values based on the magnetic pole position detection value θ1. v _u ^* , v _v ^* , v _w ^* .

The rectifier circuit 3a rectifies the three-phase AC voltage from the three-phase AC power supply 4a, converts it into a DC voltage, and supplies this DC voltage to the power converter 2a such as an inverter.

PWM (Pulse Width Modulation) circuit 19a adjusts the output voltage of power converter 2a to phase voltage command values vu ^{*, vv*, vw*} _based ^on _phase ^voltage command values _vu ^{*, vv*} _{, vw *} ^. Generate a plurality of gating signals to control _w ^* . Based on a plurality of gate signals from the PWM circuit 19a, the power converter 2a controls a plurality of semiconductor switching elements inside the power converter 2a to adjust the terminal voltage of the synchronous motor 1a such as SynRM to the phase voltage command value. Control to v _u ^* , v _v ^* , v _w ^* .

(2.2) First Example of Method for Calculating K _Ld and K _sd A first example of method for calculating K _Ld and K _sd in the first embodiment will be described.

For SynRM having the magnetic flux model formulas shown in Equations 3 and 4, let us consider a case where a sinusoidal alternating voltage is applied to the d-axis to generate magnetic fluxes as shown in Equations 9 and 10.

In this case, a current containing a DC component and harmonics shown in Equation 11 flows through the d-axis. Also, the q-axis current becomes zero as shown in Equation (12).

Since the q-axis current is zero from Equation 12, Equation 13 is obtained by setting i _q in Equation 3 to zero.

According to Equation 13, it is possible to calculate K _Ld and K _sd by setting up and solving simultaneous equations including two sets of Ψ _d and _id . Therefore, two patterns of d-axis current are generated as shown in measurement numbers 1 and 2 shown in FIG. At that time, since the magnetic flux cannot be observed directly, it must be calculated from the induced voltage, but since the induced voltage due to the magnetic flux DC component does not occur when the rotor is stationary, Ψ _d(0) in Equation 9 is cannot be calculated. Therefore, Ψ _d(0) is eliminated by the following procedure.

Rewriting Equation 3 into a function of the alternating current phase _θh using Equations 9 and 11 yields Equations 14-16. However, by setting _id to a value of zero or more, the absolute value in Expression 3 is ignored.

In order to eliminate Ψ _{d (0)} in Equation 15, differentiating Equation 14 with respect to θ _h yields K _Ld , K _sd and magnetic flux and current is obtained.

However, when solving Equation 17 as it is, it is difficult to handle K _sd in the _quadratic equation _. (20) is obtained.

Here, considering the operating point of θ _h =0° where Ψ _d ′≧0 and i _d ′≧0 clearly holds, Equation 20 can be expressed as Equation 21 using Fourier coefficients.

Since the magnetic flux is in phase with the current, and the voltage drop caused by the magnetic flux leads the magnetic flux by 90°, the Fourier sine coefficient Ψ _d(b1) of the fundamental wave component of the d-axis magnetic flux is calculated by Equation 22 be done.

From Equation 21, simultaneous equations can be established as Equations 23 and 24 from the measurement results of the magnetic flux and current of measurement numbers 1 and 2. Note that the subscripts of the Fourier coefficients represent measurement numbers.

Solving Equations 23 and 24 for K _sd , K _sd can be calculated as Equation 25.

Also, K _Ld can be calculated by Equation 26 using K _sd calculated by Equation 25.

(2.3) Second Example of Method for Calculating K _Ld and K _sd A second example of method for calculating K _Ld and K _sd in the first embodiment will be described.

Consider a case where a sinusoidal alternating voltage is applied to the d-axis of SynRM having the magnetic flux model equations shown in Equations 3 and 4 to generate magnetic fluxes as shown in Equations 27 and 28.

In this case, a current including harmonics shown in Equation 29 flows through the d-axis. Also, the current on the q-axis becomes zero as shown in Equation (30).

Since the q-axis current is zero from Equation 30, Equation 31 is obtained by setting i _q in Equation 3 to zero.

According to Equation 31, it is possible to calculate K _Ld and K _sd by setting up and solving simultaneous equations including two sets of Ψ _d and _id . Therefore, two patterns of d-axis current with different amplitudes are generated as shown in measurement numbers 1 and 2 shown in FIG.

Rewriting Equation 3 into a function of the alternating current phase _θh using Equations 27 and 29 yields Equations 32-34.

Considering the operating point of θ _h =90°, Equation 32 can be expressed as Equation 35 using Fourier coefficients. Also, since i _d ≧0, the absolute value symbol in the denominator is ignored.

The magnetic _flux is in phase with the current, and the voltage drop caused by the magnetic flux is 90° ahead of the magnetic flux. be done.

From Equation 36, simultaneous equations can be established as Equations 37 and 38 from the measurement results of the magnetic flux and current of measurement numbers 1 and 2. Note that the subscripts of the Fourier coefficients represent measurement numbers.

Solving Equations 37 and 38 for K _sd , K _sd can be calculated as Equation 39.

Also, K _Ld can be calculated by Equation 40 using K _sd calculated by Equation 39.

(2.4) Third Example of Method for Calculating K _Ld and K _sd A third example of method for calculating K _Ld and K _sd in the first embodiment will be described.

Considering the case where two patterns of d-axis current are generated as in measurement numbers 1 and 2 shown in FIG. Become. Also, the magnetic flux and current generated on the d-axis in measurement number 2 are given by equations 9 and 11, respectively. From these equations and the magnetic flux measurement results of measurement numbers 1 and 2, simultaneous equations can be established as shown in equations 41 and 42.

Solving Equations 41 and 42 for K _sd , K _sd can be calculated as Equations 43-45.

Also, K _Ld can be calculated by Equation 46 using K _sd calculated by Equations 43 to 45.

(2.5) Estimation of parameters According to the above-described calculation method (that is, any one of the first to third examples of calculation methods for K _Ld and K _sd ), the synchronous motor 1a such as SynRM It is possible to calculate K _Ld and K _sd without load operation. By executing the above-described calculation by this calculation method in the parameter estimator 20a of FIG. 1, auto-tuning of the magnetic flux model of the synchronous motor 1a such as SynRM in the first embodiment is realized, and the magnetic flux model is constructed.

<3. Second and Third Embodiments>
(3.1) Overall Configuration of Second and Third Embodiments FIG. 5 is a block diagram showing a control device according to the second to fifth embodiments together with a main circuit. A method for controlling the voltage and current of a motor (hereinafter also simply referred to as a motor or SM) will be described together with the configuration of the control device. The control calculation of the voltage and current supplied to the synchronous motor by the power converter is performed on the d- and q-axis orthogonal rotating coordinates, and the magnetic pole (N-pole) direction of the rotor of the motor is the d-axis, A direction leading 90° from the d-axis parallel to the direction is defined as the q-axis.

The control device 100b of the present embodiment controls the DC component of the d-axis current to a command value, and controls the magnetic saturation characteristics of the motor iron core based on the current and voltage when a sinusoidal alternating voltage is applied to the q-axis. At least auto-tuning is performed to calculate each parameter of the considered magnetic flux model.

In FIG. 5, the integrator 5b integrates the angular frequency _ωh of the alternating voltage to calculate the angle _θh of the q-axis alternating current command value. The d-axis alternating current command calculator 6b calculates the AC component i _qh ^* of the q-axis alternating current command value as shown in Equation (47).

The current coordinate converter 8b converts the u- and w-phase current detection values i _u and i _w respectively detected by the u-phase current detector 9ub and the w-phase current detector 9wb to the magnetic pole position detection value θ ₁ of the synchronous motor 1b. coordinate conversion to the _d- and _q -axis current detection values id and iq.

The low-pass filter 10b removes high frequency components from the _d -axis current detection value id to calculate the d-axis current detection value _idf . The low-pass filter 10c removes high frequency components from the _q -axis current detection value iq to calculate the q-axis current detection value _iqf .

The deviation between the d-axis current command value _id(0) ^* and the d-axis current detection value _idf is calculated by the subtractor 11b, and this deviation is amplified by the d-axis current controller 12b to obtain the d-axis voltage command value v _dACR (d-axis voltage feedback control value v _dACR ) is calculated. The d-axis current controller 12b operates so that the deviation between the d-axis current command value id ₍₀₎ ^* and the d-axis current detection value _{idf becomes zero to calculate the d-axis voltage feedback control value vdACR .} . Further, the deviation between the q-axis current command value i _qh ^* and the q-axis current detection value i _qf is calculated by the subtractor 11c, and this deviation is amplified by the q-axis current controller 12c to obtain the q-axis voltage command value _vqACR. (q-axis voltage feedback control value v _qACR ) is calculated. The q-axis current controller 12c operates so that the deviation between the q-axis current command value _iqh ^* and the q-axis current detection value _iqf becomes zero to calculate the q-axis voltage feedback control value _vqACR .

As will be described later, the parameters of the magnetic flux model are estimated using harmonic currents that flow due to magnetic saturation of the motor core. Therefore, in order to prevent the d-axis current regulator 12b and the q-axis current regulator 12c from acting on harmonic currents, the response frequencies of the regulators 12b and 12c and the cutoff frequencies of the low-pass filters 10b and 10c are It is set smaller than twice the angular frequency _ωh of the alternating voltage.

The DC armature resistance compensator 13b calculates the d-axis armature resistance feedforward compensation value _vdra by Equation 6, as in the first embodiment.

The voltage compensation value calculator 14b uses the q-axis reactance estimated value _Xqhest and the q-axis armature resistance estimated value _Rqhest calculated by the impedance estimator 15b to calculate the q-axis voltage feedforward compensation value _{vqhFF according} to Equation 48. do. Incidentally, the estimation of impedance such as the q-axis reactance estimated value X _qhest and the q-axis armature resistance estimated value R _qhest may be performed by the impedance estimating unit 15b using the known technique described in Patent Document 1, for example. It is possible.

The Fourier coefficient calculator 16b calculates Fourier coefficients of the q-axis voltage, _q -axis current, and _d -axis current based on _θh , vq ^* , and id. The Fourier coefficient calculator 16b calculates the Fourier cosine coefficient _Vq(a1) of the fundamental wave component of the q-axis voltage, the Fourier sine coefficient _Vq(b1) of the fundamental wave component of the q-axis voltage, and the Fourier coefficient of the fundamental wave component of the q-axis current. Cosine coefficient I _q(a1) Fourier sine coefficient I _q(b1) of fundamental wave component of q-axis current Fourier cosine coefficient I _q(a3) of triple harmonic component of q-axis current Three times q-axis current Fourier sine coefficient _Iq(b3) of the harmonic component, Fourier coefficient _Id(0) of the DC component of the d-axis current, Fourier cosine coefficient _Id(a2) of the double harmonic component of the d-axis current, and d Calculate the Fourier sine coefficient _Id(b2) of the double harmonic component of the shaft current. Fourier coefficients can be calculated using a known technique described in Patent Document 1, for example.

The Fourier cosine coefficient I _d(a2) of the double harmonic component of the d-axis current, the Fourier sine coefficient I _d(b2) of the double harmonic component of the d-axis current, and the q-axis calculated by the Fourier coefficient calculator 16b. From the Fourier cosine coefficient _Iq(a3) of the third harmonic component of the current and the Fourier sine coefficient _Iq(b3) of the third harmonic component of the q-axis current, the armature resistance compensator 17b is calculated by Equation 49 as follows: A harmonic armature resistance compensation value _vdhra and a q-axis harmonic armature resistance compensation value _Vqhra are calculated.

The _d -axis voltage command value vd ^* is obtained by adding the d-axis voltage feedback control value _vdACR , the d-axis armature resistance feedforward compensation value _vdra , and the d-axis harmonic armature resistance compensation value _vdhra by the adder 7d. calculated as On the other hand, the q-axis voltage command value vq ^* is obtained by adding the _q -axis voltage feedback control value _vqACR , the q-axis voltage feedforward compensation value _qhFF , and the q-axis harmonic armature resistance compensation value _vqhra by adders 7e and 7f. calculated as

The _d- and _q -axis voltage command values vd ^* and vq ^* calculated as described above are converted by the voltage coordinate converter 18b to the u-, v- and w-phase _phase voltage command values based on the magnetic pole position detection value θ1. v _u ^* , v _v ^* , v _w ^* .

The rectifier circuit 3b rectifies the three-phase AC voltage from the three-phase AC power supply 4b, converts it into a DC voltage, and supplies this DC voltage to the power converter 2b such as an inverter.

The PWM circuit 19b controls the output voltage of the power converter 2b to the phase voltage command values _{vu*, vv*, vw} ^* _based ^on _the ^phase voltage command values _vu ^* , _vv ^* , _vw ^* . Generate multiple gate signals for Based on a plurality of gate signals from the PWM circuit 19b, the power converter 2b controls a plurality of semiconductor switching elements inside the power converter 2b to change the terminal voltage of the synchronous motor 1b such as IPMSM or SynRM to a phase voltage. Control to command values _vu ^* , _vv ^* , _vw ^* .

(3.2) Method of Calculating K _sdq , _KLq , K _sq , and K _sqd For the magnetic flux model formulas 3 and 4, K _sdq , _KLq , K _sq , and K _sqd are calculated without no-load operation. The method is shown below.

For SynRM having the magnetic flux model formulas shown in Equations 3 and 4, consider a case where a DC voltage is applied to the d-axis and a sinusoidal alternating voltage is applied to the q-axis to generate magnetic fluxes as shown in Equations 50 and 51.

Also, the Fourier sine coefficient ψ _q(b1) of the fundamental wave component of the q-axis magnetic flux in Equation 51 is calculated by Equation 52.

In this case, the current containing the DC component and the harmonic component shown in Equation 53 flows through the d-axis, and the current containing the alternating frequency fundamental wave component and the harmonic component shown in Equation 54 flows through the q-axis.

(3.2.1) Calculation of K _sdq First, a method of calculating K _sdq using the calculation result of the first embodiment described above will be described. Solving Equation 3 for K _sdq yields Equation 55.

K _sdq can be calculated by substituting K _Ld and K _sd calculated in the first embodiment and the magnetic fluxes and currents of Expressions 50 to 54 into Expression 55.

A specific calculation method for K _sdq is shown below. By rewriting Equation 55 using Equations 50-54, Equations 56-58 are obtained.

Also, as described in the first embodiment, Ψ _d(0) cannot be directly measured when the rotor is stationary. Therefore, Ψ _d(0) is calculated based on K _Ld and K _sd adjusted previously. According to Equation 50, the d-axis magnetic flux does not contain an AC component, so it is constant regardless of the operating point. Therefore, if θ _h in Equations 53 and 54 is zero (θ _h =0°), the d-axis current and q-axis current are expressed by Equations 59 and 60, respectively.

Ψ _d(0) can be calculated by Equation 61 from Equations 3, 59, and 60.

K _sdq can be calculated by Equation 62 by setting the operating point of Equation 56 to θ _h =90° and using Ψ _d(0) calculated by Equation 61.

(3.2.2) Calculation of K _Lq , K _sq , and K _sqd For SynRM having the magnetic flux model formula shown in Formula 4, magnetic fluxes shown in Formulas 50 and 51 are generated as in the case of the calculation of K _sdq . The d-axis current command value at that time changes the DC component as indicated by measurement numbers 1 to 6 in FIG. In the current flowing at this time, harmonic components are generated as in Equations 53 and 54. For this current, the _Fourier coefficient is measured for each measurement number in FIG. Calculate and sample every time.

Next, a method of calculating K _Lq , K _sq , and K _sqd from sampled data will be described using the method of least squares, which is a system identification method, as an example. The method of least squares is a method of determining parameters so as to minimize the sum of the squared errors between the actual measured values and the values calculated by the model for all sample data.

Assuming that K _Lq , K _sq , and K _sqd are parameters for identification, Sn _q , which is the sum of the squared errors in Equation 4, is expressed by Equation 63. Note that the suffix n in the formula means the sample point number.

K _Lq , K _sq and K _sqd can be identified by transitioning K _Lq , K _sq and K _sqd so that Sn _q in Equation 63 is minimized. The parameter transition direction for decreasing Sn _q can be known by considering Sn _q as a vector having K _Lq , K _sq , and K _sqd as elements and calculating the gradient vector ∇Sn _q . Assuming that i _dn and i _qn are all positive, Equations 64 to 68 represent expressions for ∇Sn _q .

Gradient vectors can be obtained from Equations 64 to 68, so by using a local search method such as the descent method, it is possible to calculate parameters that minimize Sn _q . In addition, the initial parameter values for searching may all be zero, or approximate values calculated from sample data may be used.

By adding the equivalent magnetizing current I ₀ as a known value to i _dn in Equations 64 to 68, the above method can also be applied to the IPMSM. Expressions 69 to 72 at this time are shown.

(3.3) Estimation of Parameters According to the above-described method for calculating K _sdq , K _Lq , K _sq and K _sqd , K _sdq and K _Lq can be calculated without performing no-load operation of the synchronous motor 1b such as SynRM or IPMSM. , K _sq , K _sqd . The parameter estimator 20b of FIG. 5 performs the above calculations according to this calculation method, thereby realizing auto-tuning of the magnetic flux model of the synchronous motor 1b in the second and third embodiments, and constructing the magnetic flux model.

<4. Fourth and Fifth Embodiments>
Regarding the method of using instantaneous values instead of the method of using the Fourier coefficients of the d- and q-axis currents shown in the second and third embodiments when calculating K _sdq , K _Lq , K _sq , and K _sqd explain. Note that the control operation for applying the alternating voltage to the synchronous motor is the same as in the second and third embodiments, so description thereof will be omitted.

(4.1) Calculation of K _sdq First, the instantaneous value of the q-axis current and the instantaneous value of the d-axis current are measured at the timing of θ _h3 in FIG. 7, for example. Also, the d-axis magnetic flux is constant with only the DC component. Therefore, using the Fourier coefficient of the d-axis current DC component and the Fourier cosine coefficient of the d-axis current double harmonic, the Fourier coefficient Ψ _d(0) of the d-axis magnetic flux DC component is calculated by Equation (61). K _sdq can be calculated by calculating Equation 55 using these values.

(4.2) Calculation of K _Lq , K _sq , and _{K sqd} Select and measure at least the number of identification parameters. The instantaneous value of the q-axis magnetic flux is calculated by Equation 51 using the Fourier sine coefficient of the q-axis magnetic flux fundamental wave calculated by Equation 52 and θ _h1 to θ _h5 in FIG.

By using the sampled data in Equations 64 to 68 in place of the sampled data created by the Fourier coefficients and _θh , the parameters can be calculated by the least-squares method.

(4.3) Estimation of Parameters According to the above-described method of calculating K _sdq , K _Lq , K _sq and K _sqd , K _sdq and K _Lq can be calculated without performing no-load operation of the synchronous motor 1b such as SynRM or IPMSM. , K _sq , K _sqd . The parameter estimator 20b of FIG. 5 performs the above calculations according to this calculation method, thereby realizing auto-tuning of the magnetic flux model of the synchronous motor 1b in the fourth and fifth embodiments, and constructing the magnetic flux model.

FIG. 8 is a diagram illustrating a hardware configuration of an arithmetic unit included in the control device; The control device includes a computing device that performs computation based on a magnetic flux model that takes into account at least the magnetic saturation characteristics of the motor core. FIG. 8 shows a microcomputer 110, which is an example of an arithmetic device. The microcomputer 110 includes a memory 121 , a CPU (Central Processing Unit) 122 , an AD (Analog to Digital) conversion section 123 , a PWM module 124 , a communication section 125 and a timer 126 . The CPU 122 is a processor that controls the control device. The communication unit 125 communicates with a host controller outside the microcomputer 110 . The timer 126 counts timer values. The memory 121 stores programs and the like. A program in the memory 121 causes the CPU 122 to operate. The function of each control block in FIGS. 1 and 5 is implemented by the CPU 122 operating according to a program readable and stored in the memory 121 .

Each control block in FIG. 1 includes, for example, adders 7a, 7b, 7c, a subtractor 11a, a d-axis alternating current command calculator 6a, a d-axis current controller 12a, a voltage coordinate converter 18a, and a current coordinate converter 8a. , a DC armature resistance compensator 13a, a low-pass filter 10a, a voltage compensation value calculator 14a, an armature resistance compensator 17a, an integrator 5a, a Fourier coefficient calculator 16a, an impedance estimator 15a, and a parameter estimator 20a.

Each control block in FIG. 5 includes, for example, adders 7d, 7e, 7f, subtractors 11b, 11c, q-axis alternating current command calculator 6b, current regulators 12b, 12c, voltage coordinate converter 18b, current coordinate converter DC armature resistance compensator 13b, low-pass filters 10b and 10c, voltage compensation value calculator 14b, armature resistance compensator 17b, integrator 5b, Fourier coefficient calculator 16b, impedance estimator 15b, and parameter estimator 20b is.

The function of each control block in FIGS. 1 and 5 can be provided by a program that causes a computer to implement each function. Also, the function of each control block can be provided by a computer-readable recording medium recording the above program, or a computer program product such as the above program. Examples of recording media that can be used include flexible disks, hard disks, optical disks, magneto-optical disks, CD-ROMs, magnetic tapes, nonvolatile memory cards, and ROMs.

Although the control device for the synchronous motor has been described above with reference to the embodiments, the present invention is not limited to the above embodiments. Various modifications and improvements such as combination or replacement with part or all of other embodiments are possible within the scope of the present invention.

1a, 1b synchronous motors 2a, 2b power converters 3a, 3b rectifier circuits 4a, 4b three-phase AC power supplies 5a, 5b integrators 6a, 6b alternating current command calculators 7a to 7f adders 8a, 8b current coordinate converters 9ua, 9wa, 9ub, 9wb Current detectors 10a, 10b, 10c Low-pass filters 11a, 11b, 11c Subtractors 12a, 12b, 12c Current controllers 13a, 13b DC armature resistance compensators 14a, 14b Voltage compensation value calculators 15a, 15b Impedance estimators 16a, 16b Fourier coefficient calculators 17a, 17b Armature resistance compensators 18a, 18b Voltage coordinate converters 19a, 19b PWM circuits 20a, 20b Parameter estimators 100a, 100b Controller 110 Microcomputer

Claims (3)

The current and voltage supplied to the synchronous motor by the power converter are controlled on a d, q orthogonal rotating coordinate system consisting of a d-axis parallel to the magnetic pole direction of the rotor of the synchronous motor and a q-axis orthogonal to the d-axis. a controller,
A computing device that performs computation based on a magnetic flux model that takes into account at least the magnetic saturation characteristics of an iron core of the motor,
The computing device is
controlling the DC component of the d-axis current of the synchronous motor to a command value;
applying a sinusoidal alternating voltage to the d-axis;
calculating the Fourier coefficient of the fundamental wave component of the d-axis current;
calculating the Fourier coefficient of the double harmonic component of the d-axis current;
calculating the Fourier coefficient of the triple harmonic component of the d-axis current;
calculating the Fourier coefficient of the fundamental wave component of the d-axis voltage of the synchronous motor;
calculating the Fourier coefficient of the fundamental wave component of the d-axis magnetic flux of the synchronous motor based on the Fourier coefficient of the fundamental wave component of the d-axis voltage and the angular frequency of the alternating voltage;
Fourier coefficient of the fundamental wave component of the d-axis current, Fourier coefficient of the double harmonic component of the d-axis current, Fourier coefficient of the triple harmonic component of the d-axis current, Fourier coefficient of the DC component of the d-axis current , and a first parameter indicating the maximum value of the slope of the d-axis magnetic flux with respect to the d-axis current, based on the Fourier coefficient of the fundamental wave component of the d-axis magnetic flux, and a second parameter indicating the degree of magnetic saturation on the d-axis 2 parameters,
A control device for a synchronous motor, wherein the magnetic flux model is configured using the first parameter and the second parameter. The current and voltage supplied to the synchronous motor by the power converter are controlled on a d, q orthogonal rotating coordinate system consisting of a d-axis parallel to the magnetic pole direction of the rotor of the synchronous motor and a q-axis orthogonal to the d-axis. a controller,
A computing device that performs computation based on a magnetic flux model that takes into account at least the magnetic saturation characteristics of an iron core of the motor,
The computing device is
controlling the DC component of the d-axis current of the synchronous motor to a command value;
applying a sinusoidal alternating voltage to the q-axis;
calculating the Fourier coefficient of the fundamental wave component of the q-axis current of the synchronous motor;
calculating the Fourier coefficient of the triple harmonic component of the q-axis current;
calculating the Fourier coefficient of the DC component of the d-axis current;
calculating the Fourier coefficient of the double harmonic component of the d-axis current;
calculating the Fourier coefficient of the DC component of the d-axis magnetic flux of the synchronous motor based on the Fourier coefficient of the DC component of the d-axis current and the Fourier coefficient of the double harmonic component of the d-axis current;
Fourier coefficient of the fundamental wave component of the q-axis current, Fourier coefficient of the triple harmonic component of the q-axis current, Fourier coefficient of the DC component of the d-axis current, Fourier coefficient of the double harmonic component of the d-axis current , based on the Fourier coefficient of the DC component of the d-axis magnetic flux, a first parameter indicating the maximum value of the slope of the d-axis magnetic flux with respect to the d-axis current, and a second parameter indicating the degree of magnetic saturation on the d-axis , calculating a third parameter indicating the degree of interference of the q-axis current with the d-axis magnetic flux,
A control device for a synchronous motor, wherein the magnetic flux model is configured using the first parameter, the second parameter and the third parameter. The current and voltage supplied to the synchronous motor by the power converter are controlled on a d, q orthogonal rotating coordinate system consisting of a d-axis parallel to the magnetic pole direction of the rotor of the synchronous motor and a q-axis orthogonal to the d-axis. a controller,
A computing device that performs computation based on a magnetic flux model that takes into account at least the magnetic saturation characteristics of an iron core of the motor,
The computing device is
controlling the DC component of the d-axis current of the synchronous motor to a command value;
applying a sinusoidal alternating voltage to the q-axis;
calculating the Fourier coefficient of the DC component of the d-axis current;
calculating the Fourier coefficient of the DC component of the d-axis magnetic flux of the synchronous motor based on the Fourier coefficient of the DC component of the d-axis current and the Fourier coefficient of the double harmonic component of the d-axis current;
the d-axis current, the q-axis current of the synchronous motor, the Fourier coefficient of the DC component of the d-axis magnetic flux, a first parameter indicating the maximum value of the slope of the d-axis magnetic flux with respect to the d-axis current, and calculating a third parameter indicating the degree of interference of the q-axis current with the d-axis magnetic flux based on the second parameter indicating the degree of magnetic saturation;
A control device for a synchronous motor, wherein the magnetic flux model is configured using the first parameter, the second parameter and the third parameter.

Images